@arcaelas/mcp
A modular MCP (Model Context Protocol) server providing AI-powered tools for image generation, text-to-speech, background removal, and image upscaling.
Installation
npm install -g @arcaelas/mcp
Usage
stdio Mode
For Claude Desktop, Cline, and other MCP clients that support stdio transport:
arcaelas-mcp --stdio
Claude Desktop Configuration (~/.config/claude/claude_desktop_config.json):
{
"mcpServers": {
"arcaelas": {
"command": "npx",
"args": ["-y", "@arcaelas/mcp", "--stdio"],
"env": {
"OPENAI_API_KEY": "sk-...",
"CLIENT_ID": "your-client-id"
}
}
}
}
HTTP/SSE Mode
For clients that support HTTP transport (Cursor, etc.):
# Default port (3100)
arcaelas-mcp
# Custom port
arcaelas-mcp --port 8080
Endpoints:
| Endpoint | Method | Description |
|---|---|---|
/sse |
GET | Server-Sent Events connection |
/messages |
POST | Send messages (requires sessionId query param) |
/health |
GET | Health check and server info |
Tools
image
Generate images using AI models.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
prompt |
string | Yes | - | Image description in English |
model |
string | No | nanobanana |
Model: gptimage, nanobanana, nanobanana-pro |
size |
string | No | 1024x1024 |
Dimensions: 512x512, 1024x1024, 1024x1792, 1792x1024 |
audio
Convert text to speech with natural-sounding voices.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
text |
string | Yes | - | Text to synthesize |
voice |
string | No | nova |
Voice: nova, alloy, echo, fable, onyx, shimmer, coral, sage |
bgcleaner
Remove backgrounds from images using AI with high precision edge detection.
| Parameter | Type | Required | Description |
|---|---|---|---|
image_path |
string | Yes | Absolute path to source image (PNG, JPG, WEBP) |
Output: Returns path to folder containing processed PNG with transparent background.
resize
Upscale images using AI without quality loss.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
image_path |
string | Yes | - | Absolute path to source image |
scale |
number | No | 2 |
Scale factor: 2, 3, or 4 |
model |
string | No | plus |
Model: diffuser, plus, general |
face_enhance |
boolean | No | false |
Enhance faces (plus/general only) |
Models:
diffuser— Best for small images, more creative resultsplus— Balanced for medium to large imagesgeneral— Optimized for very large images
Environment Variables
| Variable | Required | Description |
|---|---|---|
OPENAI_API_KEY |
Yes* | OpenAI API key (required for image and audio tools) |
OPENAI_BASE_URL |
No | Custom OpenAI-compatible API endpoint |
CLIENT_ID |
Yes* | Client ID for image processing services (required for bgcleaner and resize) |
Examples
# Start HTTP server
OPENAI_API_KEY=sk-xxx CLIENT_ID=xxx arcaelas-mcp --port 3100
# Start in stdio mode
OPENAI_API_KEY=sk-xxx CLIENT_ID=xxx arcaelas-mcp --stdio
# Using npx without global install
npx -y @arcaelas/mcp --stdio
License
MIT
